A local AI-powered research paper summarizer.
Paste any research paper URL and receive a clean, structured Markdown summary including title, authors, methodology, technologies, references, research gaps, and key findings, all generated locally using Ollama.
This project extracts the full content of a research paper from a URL using the Jina AI Reader API and sends it to a locally running LLM via Ollama. The model processes the content and returns a well-structured Markdown summary.
- A research paper URL is passed to the Jina AI Reader API.
- The full page content is extracted using Jina ai.
- The extracted content is sent to a local
llama3.2:3bmodel running via Ollama. - The model generates a structured Markdown summary.
The generated output includes:
- Title
- Abstract
- Introduction
- Methods
- Key Technologies
- Results
- Discussion
- Research Gaps
- References
- Structured comparison tables (if applicable)
- LLM: llama3.2:3b via Ollama
- Web Extraction: Jina AI Reader API
- Environment: Python
- Interface: Jupyter Notebook
- Rendering: IPython.display.Markdown
- Support direct PDF upload
- Add batch processing of multiple URLs
- Improved speed
- Export summaries to PDF
- Add a simple web interface
For installation instructions and additional information, visit my contribution here.
Developed by Karam Sayed
MIT License